US20030110423A1 - Variable maximum die temperature based on performance state - Google Patents

Variable maximum die temperature based on performance state Download PDF

Info

Publication number
US20030110423A1
US20030110423A1 US10/015,031 US1503101A US2003110423A1 US 20030110423 A1 US20030110423 A1 US 20030110423A1 US 1503101 A US1503101 A US 1503101A US 2003110423 A1 US2003110423 A1 US 2003110423A1
Authority
US
United States
Prior art keywords
temperature
performance
performance state
maximum
state
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
US10/015,031
Other versions
US6889332B2 (en
Inventor
Frank Helms
Jeffrey Brinkley
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Advanced Micro Devices Inc
Original Assignee
Advanced Micro Devices Inc
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Family has litigation
First worldwide family litigation filed litigation Critical https://patents.darts-ip.com/?family=21769165&utm_source=google_patent&utm_medium=platform_link&utm_campaign=public_patent_search&patent=US20030110423(A1) "Global patent litigation dataset” by Darts-ip is licensed under a Creative Commons Attribution 4.0 International License.
Application filed by Advanced Micro Devices Inc filed Critical Advanced Micro Devices Inc
Priority to US10/015,031 priority Critical patent/US6889332B2/en
Assigned to ADVANCED MICRO DEVICES, INC. reassignment ADVANCED MICRO DEVICES, INC. ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS). Assignors: BRINKLEY, JEFFREY A., HELMS, FRANK P.
Publication of US20030110423A1 publication Critical patent/US20030110423A1/en
Application granted granted Critical
Publication of US6889332B2 publication Critical patent/US6889332B2/en
Adjusted expiration legal-status Critical
Expired - Lifetime legal-status Critical Current

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F1/00Details not covered by groups G06F3/00 - G06F13/00 and G06F21/00
    • G06F1/16Constructional details or arrangements
    • G06F1/20Cooling means
    • G06F1/206Cooling means comprising thermal management
    • YGENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02DCLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D10/00Energy efficient computing, e.g. low power processors, power management or thermal management

Definitions

  • This invention relates to computer systems and more particularly to power management of computer systems.
  • the rated processor speed can be limited by power and thermal considerations and more specifically is determined, at least in significant part, by operating frequency, voltage, and temperature. If the die temperature can be kept sufficiently low, the rated processor clock speed can be increased. In one prior art approach to maximizing frequency at which a processor can be marketed, the processor includes an on-die temperature sensor that automatically throttles (reduces) the CPU's clock by 50% if the processor's temperature crosses a factory set threshold. If the maximum die temperature is set at a sufficiently low level of, e.g., 70 degrees C., the processor is able to meet the timing budget for the silicon process in which it is manufactured at a higher frequency at a given voltage than if the die temperature was significantly higher.
  • the processor includes an on-die temperature sensor that automatically throttles (reduces) the CPU's clock by 50% if the processor's temperature crosses a factory set threshold. If the maximum die temperature is set at a sufficiently low level of, e.g., 70 degrees C., the processor is able to meet the timing
  • the timing budget is the amount of time a signal has to propagate through combinational logic from one set of storage elements (e.g., flip-flops) to another set of storage elements and meet the setup and hold times associated with the flip-flops.
  • a device can more easily meet a timing budget because, e.g., the propagation delays are reduced.
  • the processor clock is throttled back to, e.g., 50% of its rated speed, which reduces the power consumption and thus the temperature.
  • a processor rated at 2 GHz would run at 1 GHz upon crossing of the maximum die temperature threshold. In that way the processor can be rated at a higher speed at or below 70 degrees C., although it may operate with significantly reduced capability above that temperature.
  • the invention provides a system in which the maximum performance state, which is typically based on such factors as operating frequency and voltage, changes according to thermal criteria.
  • the thermal criteria may be, for example, a die temperature of the microprocessor. Once the temperature crosses above a predetermined temperature threshold, the maximum performance state available to the system is changed to a lower performance state. As the temperature crossed above successive temperature thresholds, successively lower maximum performance states become available to the system. That allows processor performance to be impacted less than the throttling alternative described above as the temperature increases. When the temperature crosses below a predetermined temperature threshold, a higher maximum performance state is again available to the system. Hysteresis may be implemented to ensure the system does not rapidly switch between performance states when the temperature is close to predetermined thresholds.
  • the invention provides a processor that has multiple performance states determined by such factors as operating frequency and/or voltage. In a first temperature range, the processor operates with a first maximum performance state and in a second temperature range, higher than the first temperature range, the processor operates with a lower maximum performance state.
  • the invention provides a method of operating a computing system that includes determining a temperature associated with an integrated circuit and operating the integrated circuit with a first performance state as a maximum performance state according to the determined temperature, the first performance state being one of a plurality of performance states available at the determined temperature.
  • a method for providing a variable maximum die temperature for an integrated circuit in a computer system having a plurality of different maximum performance states comprising increasing the maximum allowable die temperature as a maximum allowable performance state is decreased.
  • FIG. 1 is a high level flow diagram of power management operations based on CPU utilization.
  • FIG. 2 illustrates switching between performance states according to utilization.
  • FIG. 3 illustrates an exemplary processing system that can change performance state.
  • FIG. 4 illustrates a table including various temperature trip points for an exemplary system.
  • FIG. 5 shows the high level flow of an embodiment of the invention in which the maximum performance state available is adjusted according to temperature measurements.
  • a computer system has a plurality of processor performance states, the processor performance states being generally determined by unique voltage/frequency combinations.
  • the computing system provides lower maximum performance states as the detected temperature of, e.g. the processor die, increases.
  • the computer system may change to a different performance states based on such factors as processor utilization as explained more fully herein.
  • the detected temperatures may be related directly to the processor by measuring the die temperature directly using a sensor on the die or may measure the die temperature indirectly using a sensor that is off-die.
  • other indirect measurements may be taken such as ambient temperatures or measurements of the temperature of other computer system components to determine the appropriate mode of operation although indirect measurements may reduce accuracy. Because the different temperature ranges have different maximum performance states, the maximum amount of power that can be consumed grows less as the temperature increases through the various temperature ranges and lower maximum performance states are provided.
  • processor performance states could be determined by an amount of chip area that was enabled for processing.
  • portions of the processor may be selectively disabled to control the amount of power consumed.
  • the number of execution units that are enabled may vary.
  • a power management control function in the computer system periodically determines the utilization level of the processor, i.e., how much of the available processor resources are being utilized, and selects a performance state that is appropriate for the utilization level.
  • FIG. 1 a flow diagram illustrates at a high level, operation of an embodiment of a power management function utilized to provide the requisite power management control.
  • the current CPU utilization is periodically determined in 101 . That current utilization is then compared to a high threshold level, e.g., 80% of processing resources, in 103 .
  • the power management control function increases the performance state of the processor in 105 . In one embodiment, that can be accomplished by selecting a voltage/frequency pair that provides greater performance and then causing the processor to operate at the new voltage and frequency, as described further herein.
  • the current utilization is below the high threshold, then the current utilization is compared to the low threshold in 107 .
  • An exemplary low threshold level is 55%. If the current utilization is below that low threshold, the power management control function decreases the performance state of the processor in 109 . As described further herein, that may be accomplished by selecting, e.g., a voltage/frequency pair providing lower performance and then causing the performance change to occur. The power management control function then returns to 101 to periodically determine the utilization level and again compare the current utilization level to the high and low threshold levels. In that way the power management can tailor the performance state of the processor to the actual requirements.
  • the management control function determines that more performance is necessary to meet performance requirements, one approach to providing increased performance is to increase the performance one step at a time until the current utilization is below the high threshold level.
  • the power management control function selects the highest performance state regardless of the current performance state.
  • the reasons for always selecting the highest possible performance state when a higher performance state is needed are as follows. In computer systems, performance demands are often of a bursty nature. When a higher performance state is required based on the current utilization level, stepping the performance state to a next higher level can result in degradation of performance that can be perceived by the user. That is especially true when the task that needs the increased performance requires a near real-time response, for instance, while decoding an audio or video file.
  • FIG. 2 illustrates that concept. Assume the processor has five performance states P 0 -P 4 , with P 0 being the highest and P 4 the lowest. Whenever the power management determines that a higher performance state is required when operating at any of the levels P 1 -P 4 , the power management selects the maximum performance state P 0 as the next performance state. Thus, in one embodiment, if the performance state is always taken straight to the maximum performance state when a performance increase is required, rather than stepping up to the maximum performance state, there is less of a chance that a user could notice any performance degradation. In effect, the power management control function anticipates a peak loading by assuming that any indication of a required increase in performance is assumed to be a burst requiring peak performance.
  • next lower performance state is selected.
  • P 1 is selected as the next lower performance state.
  • P 2 when a performance decrease is effectuated by the power management control function. In that way, if the performance is still too high, successively lower performance states can be selected and the chance than any degradation is detected by a system user is reduced.
  • the power management control function selects the maximum (or near maximum) performance state, while a decrease in performance causes the power management control function to step to the next lowest performance state.
  • processor utilization is kept within the range of the high and low thresholds, then a user should experience a crisp, high performance system, while still getting the benefit of power savings for those applications or those portions of applications requiring less performance. That approach reduces power consumption, extends battery life, reduces temperature resulting in less need for cooling and thus less fan noise, while still maintaining high performance and thus maintaining a perception of fast response to the user. Note that running at a lower average CPU die temperature increases CPU reliability, and that a lower CPU temperature results in a lower system temperature, which increases system reliability.
  • the power management software has to cause the voltage and frequency used by the CPU to change.
  • a processor is shown that can dynamically adjust its frequency and/or operating voltage to provide better thermal and power management in accordance with processor utilization.
  • Processor 301 includes a programmable voltage control field 303 , core clock frequency control field 304 and count field 305 . Those fields may be located in one or more programmable registers.
  • the processor and/or system determines that a change to the operating voltage and/or frequency is desired to increase or decrease the performance state, the desired frequency and voltage control information are loaded into frequency control field 304 and voltage control field 303 , respectively.
  • Access to a register containing those fields, or an access to another register location, or access to a particular field in one of those registers can be used as a trigger to initiate the protocol for a performance state change.
  • the protocol which will vary according to processor design, ensures that the processor enters a quiescent state that is suitable for performance state transitions.
  • bus interface 310 is implemented separately from CPU core logic 311 and functions to hold off traffic to CPU core logic 311 during performance state transitions.
  • Various messages may be sent or signals provided indicating that the processor has or is about to enter the quiescent state.
  • various control and/or status signals 309 and 313 may be provided to and from core logic 311 .
  • various power planes and clock grids may ensure some parts of the processor are unaffected by the processor core logic 311 entering the quiescent state.
  • the quiescent state may vary according to processor design. For example, for one processor the quiescent state may have the clocks turned off and in other processors, the clocks may be reduced to an extent to allow appropriate voltage changes.
  • the power state changes may include clock frequency changes, voltage changes or both. In x86 architectures, such a quiescent state has been referred to as the “stop grant” state in which execution of operating system and application code is stopped.
  • stop grant state in which execution of operating system and application code is stopped.
  • other ways may be used to initiate the protocol necessary to enter the quiescent state suitable for performance state changes. For example, initiation of the protocol to enter the quiescent state may be indicated by writing a command over a communication link coupling the processor to another device.
  • the processor can transition the voltage and frequency to the new states specified in voltage control field 303 and clock frequency control field 304 .
  • the processor core clocks are stopped after the processor enters the quiescent state to suitable for changing performance states.
  • the processor core clock frequency is reduced to a frequency which can safely tolerate desired voltage changes.
  • clock control frequency information is supplied as multiplier values for a clock that is supplied to processor 301 .
  • the voltage control information specified in voltage control field 303 is supplied to voltage regulator 315 which in turn supplies CPU core logic 311 with the new voltage during the quiescent state.
  • the voltage information may be supplied from a device outside the processor, or may be supplied by a communication link with a voltage regulator circuit.
  • the quiescent state is typically maintained for a period of time to allow the new voltage and clock frequency to stabilize. That time period may be specified in count field 305 and counted using a counter in the processor or in another location in the computer system. Once the count is complete, the protocol is initiated to exit the quiescent state and resume executing operating system and application code, including restarting any buses that have been “turned off” by the processor as part of entering the quiescent state.
  • FIG. 4 illustrates an exemplary set of temperature “trip points” at which power management activities take place.
  • the power management control described herein may be implemented as part of the operating system (OS), in whole or in part.
  • separate power management applications may incorporate some or all of the capability described herein.
  • the various power management tasks may be shared between the operating system, software applications running under the operating system and BIOS routines.
  • the power management techniques may be implemented in hardware.
  • a temperature trip point that causes the system to make a higher maximum performance state available differs from the temperature trip point that causes a lower maximum performance state to be available. That provides hysteresis and avoids frequent changes in the available maximum performance state if the temperature is at or near a trip point.
  • hysteresis is implemented by causing P 0 to become unavailable as the maximum performance state at 70C but to again become available at a temperature trip point of 68C.
  • P 0 becomes the maximum performance state.
  • P 1 becomes unavailable as the maximum performance state at 80C but again become available as the maximum performance state when the temperature crosses back below 78C.
  • the system may detect whether the temperature is greater than, greater than or equal to, equal to, less than or less than or equal to a particular temperature, as appropriate for the particular situation, to determine if the system has crossed a temperature trip point.
  • the available performance states may be implemented as a table or other data structure in BIOS, in application software, or in a power management portion of the operating system software.
  • Other ways to encode the appropriate performance states according to the temperature would be readily apparent to those of skill in the art.
  • FIG. 5 provides a high level illustration of the operation of an embodiment of the invention.
  • FIG. 5 deals with maximum performance state transitions and does not illustrate operations of all power management functions.
  • the flow chart corresponds to an embodiment of the invention that utilizes three temperature ranges illustrated in FIG. 4.
  • a temperature associated with the microprocessor e.g., a die temperature of the microprocessor
  • the system checks to see if it is operating with the maximum performance state being P 0 in 503 . If so, the system checks to see if the temperature has crossed above the 70C trip point. If so, at 507 , the system changes the maximum available performance state to P 1 and returns to obtain another temperature reading.
  • the system checks to see at 509 if the maximum performance state is P 1 . If so, system checks to see if the temperature has crossed below 68C in 511 . If so, the maximum performance state is changed to P 0 in 513 . If not, the system checks in 515 if the temperature has crossed above the 80C trip point. If so, the maximum performance state is reduced to P 2 in 517 .
  • the system determines that the maximum performance state is P 2 and the system checks to see if the temperature is less than or equal to the 78C trip point in 521 . If so, the maximum performance state is increased to P 1 in 523 and the system returns to 500 to obtain a current temperature.
  • FIG. 5 provides for increasing or decreasing the maximum performance state by one performance state.
  • the system software and/or hardware may allow for maximum performance state changes of more than one performance state.
  • the number of trip points can of course vary depending on system design.
  • hysteresis is implemented at 511 and 521 such that the temperature at which the system switches back to a higher maximum performance state is different (and lower) than the temperature at which it switched away from that maximum performance state to a lower maximum performance state. While implementation of hysteresis is not necessary to practice the invention, it does prevent frequent switching between maximum allowable performance states when the temperature is hovering around a trip point.
  • the maximum performance state remains the same. The process continues periodically evaluating temperatures and setting the appropriate maximum performance state based on the evaluated temperature. In that way, the maximum performance state is determined based on thermal criteria. The system is free to switch to lower performance states according to power management considerations such as processor utilization.
  • determining when temperature thresholds have been crossed may be interrupt driven.
  • temperature sensor apparatus may include programmable high and low thresholds, and generate an interrupt when one of the thresholds is crossed. If the threshold that is crossed causes a maximum performance state change, the high and low thresholds may be set to new threshold values appropriate for the new maximum performance state. The control software is then invoked to change maximum performance states on occurrence of another interrupt.
  • a computer dynamically adapts its maximum performance state according to a thermal indicator.
  • the description of the invention set forth herein is illustrative, and is not intended to limit the scope of the invention as set forth in the following claims.
  • the teachings herein may also be utilized in other computing devices, such as servers, multi-processor systems, work stations and/or portable computing devices such as personal digital assistants, (PDAs), which are handheld devices that typically combine computing, telephone/fax, and networking features or in other computing and/or communication equipment, where such power management techniques described herein may prove useful.
  • PDAs personal digital assistants
  • Other variations and modifications of the embodiments disclosed herein may be made based on the description set forth herein, without departing from the scope and spirit of the invention as set forth in the following claims.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Human Computer Interaction (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Power Sources (AREA)

Abstract

The maximum performance state available to a processor in a computer system, in terms of operating frequency and/or voltage, changes according to thermal criteria. When the temperature increases above a predetermined threshold, the maximum performance state available is reduced. Multiple temperature thresholds may be utilized providing for a gradually reduced maximum performance state as temperature increases. When the temperature returns to a lower level, the maximum performance state available is increased. Changing the maximum available performance state according to temperature provides for more gradual reduction in performance as temperature increases, which results in higher average system performance as temperature increases. Thus, a more gradual reduction in performance is provided while still maintaining a high speed rating of the processor in more ideal conditions. In normal operating conditions, high processor performance is provided, while slightly reduced performance is provided in abnormal operating conditions.

Description

    BACKGROUND
  • 1. Field of the Invention [0001]
  • This invention relates to computer systems and more particularly to power management of computer systems. [0002]
  • 2. Description of the Related Art [0003]
  • An important market criteria for microprocessors is the rated frequency at which the microprocessor operates, typically denominated in hundreds of MHz or GHz. Rated frequency is used as a surrogate for performance and higher performance processors command higher prices. Thus, it is advantageous to market processors with the highest possible frequency rating. [0004]
  • However, there are limitations as to how high one may rate a processor in terms of frequency. As the number of transistors provided on a single integrated circuit has increased and the clock speed of integrated circuits has increased, power and related thermal considerations have become an important consideration in computer design. Power and thermal considerations can limit processor performance in certain environments, particularly in mobile applications. Computer systems measure the die temperature or case temperature using sensors on or close to the processor. In order to keep processor die temperatures within safe bounds to avoid potential damage to the processor, passive and active cooling have traditionally been employed to control temperature. Passive cooling has been accomplished by reducing a processors clocks speed, either by throttling processor clocks (reducing effective frequency by turning clocks off for a predetermined period) or by reducing actual frequency of the clocks. Reducing the actual and/or effective frequency of an integrated circuit causes a reduction in power consumption and thus reduces temperature. In addition to reducing clock frequency, it is known in the art to reduce voltage in conjunction with reduced clock speed to achieve additional cooling. [0005]
  • Thus, the rated processor speed can be limited by power and thermal considerations and more specifically is determined, at least in significant part, by operating frequency, voltage, and temperature. If the die temperature can be kept sufficiently low, the rated processor clock speed can be increased. In one prior art approach to maximizing frequency at which a processor can be marketed, the processor includes an on-die temperature sensor that automatically throttles (reduces) the CPU's clock by 50% if the processor's temperature crosses a factory set threshold. If the maximum die temperature is set at a sufficiently low level of, e.g., 70 degrees C., the processor is able to meet the timing budget for the silicon process in which it is manufactured at a higher frequency at a given voltage than if the die temperature was significantly higher. The timing budget is the amount of time a signal has to propagate through combinational logic from one set of storage elements (e.g., flip-flops) to another set of storage elements and meet the setup and hold times associated with the flip-flops. At a lower temperature, a device can more easily meet a timing budget because, e.g., the propagation delays are reduced. Thus, to ensure that timing budgets are met, in the prior art approach described above, when the maximum die temperature associated with the device is crossed, e.g., 70 degrees C., the processor clock is throttled back to, e.g., 50% of its rated speed, which reduces the power consumption and thus the temperature. Thus, a processor rated at 2 GHz would run at 1 GHz upon crossing of the maximum die temperature threshold. In that way the processor can be rated at a higher speed at or below 70 degrees C., although it may operate with significantly reduced capability above that temperature. [0006]
  • While that allows the processor to be rated highly, it also results in a system that operates at a significantly lower clock speed in certain environments. It would be desirable to ship more higher speed rated processors without having to drastically sacrifice performance when the critical temperature threshold is crossed. [0007]
  • SUMMARY
  • Accordingly, the invention provides a system in which the maximum performance state, which is typically based on such factors as operating frequency and voltage, changes according to thermal criteria. The thermal criteria may be, for example, a die temperature of the microprocessor. Once the temperature crosses above a predetermined temperature threshold, the maximum performance state available to the system is changed to a lower performance state. As the temperature crossed above successive temperature thresholds, successively lower maximum performance states become available to the system. That allows processor performance to be impacted less than the throttling alternative described above as the temperature increases. When the temperature crosses below a predetermined temperature threshold, a higher maximum performance state is again available to the system. Hysteresis may be implemented to ensure the system does not rapidly switch between performance states when the temperature is close to predetermined thresholds. [0008]
  • In an embodiment, the invention provides a processor that has multiple performance states determined by such factors as operating frequency and/or voltage. In a first temperature range, the processor operates with a first maximum performance state and in a second temperature range, higher than the first temperature range, the processor operates with a lower maximum performance state. [0009]
  • In another embodiment the invention provides a method of operating a computing system that includes determining a temperature associated with an integrated circuit and operating the integrated circuit with a first performance state as a maximum performance state according to the determined temperature, the first performance state being one of a plurality of performance states available at the determined temperature. [0010]
  • In another embodiment a method for providing a variable maximum die temperature for an integrated circuit in a computer system having a plurality of different maximum performance states, comprising increasing the maximum allowable die temperature as a maximum allowable performance state is decreased. [0011]
  • BRIEF DESCRIPTION OF THE DRAWINGS
  • The present invention may be better understood, and its numerous objects, features, and advantages made apparent to those skilled in the art by referencing the accompanying drawings in which the use of the same reference symbols in different drawings indicates similar or identical items. [0012]
  • FIG. 1 is a high level flow diagram of power management operations based on CPU utilization. [0013]
  • FIG. 2 illustrates switching between performance states according to utilization. [0014]
  • FIG. 3 illustrates an exemplary processing system that can change performance state. [0015]
  • FIG. 4 illustrates a table including various temperature trip points for an exemplary system. [0016]
  • FIG. 5 shows the high level flow of an embodiment of the invention in which the maximum performance state available is adjusted according to temperature measurements.[0017]
  • DESCRIPTION OF THE PREFERRED EMBODIMENTS
  • A computer system according to one embodiment of the invention has a plurality of processor performance states, the processor performance states being generally determined by unique voltage/frequency combinations. The computing system provides lower maximum performance states as the detected temperature of, e.g. the processor die, increases. In addition to having a different maximum performance state associated with particular temperature ranges, the computer system may change to a different performance states based on such factors as processor utilization as explained more fully herein. The detected temperatures may be related directly to the processor by measuring the die temperature directly using a sensor on the die or may measure the die temperature indirectly using a sensor that is off-die. In addition, other indirect measurements may be taken such as ambient temperatures or measurements of the temperature of other computer system components to determine the appropriate mode of operation although indirect measurements may reduce accuracy. Because the different temperature ranges have different maximum performance states, the maximum amount of power that can be consumed grows less as the temperature increases through the various temperature ranges and lower maximum performance states are provided. [0018]
  • While a performance state is commonly based in operating voltage and frequency, other criteria can also be used. For example, in some embodiments, processor performance states could be determined by an amount of chip area that was enabled for processing. Thus, portions of the processor may be selectively disabled to control the amount of power consumed. For example, the number of execution units that are enabled may vary. [0019]
  • In order to more fully appreciate the invention, the use of performance states is described in which a power management control function in the computer system periodically determines the utilization level of the processor, i.e., how much of the available processor resources are being utilized, and selects a performance state that is appropriate for the utilization level. Referring to FIG. 1, a flow diagram illustrates at a high level, operation of an embodiment of a power management function utilized to provide the requisite power management control. The current CPU utilization is periodically determined in [0020] 101. That current utilization is then compared to a high threshold level, e.g., 80% of processing resources, in 103. If the utilization level is above the high threshold level, indicating that processor resources are being utilized at a level above 80%, the power management control function increases the performance state of the processor in 105. In one embodiment, that can be accomplished by selecting a voltage/frequency pair that provides greater performance and then causing the processor to operate at the new voltage and frequency, as described further herein.
  • If the current utilization is below the high threshold, then the current utilization is compared to the low threshold in [0021] 107. An exemplary low threshold level is 55%. If the current utilization is below that low threshold, the power management control function decreases the performance state of the processor in 109. As described further herein, that may be accomplished by selecting, e.g., a voltage/frequency pair providing lower performance and then causing the performance change to occur. The power management control function then returns to 101 to periodically determine the utilization level and again compare the current utilization level to the high and low threshold levels. In that way the power management can tailor the performance state of the processor to the actual requirements.
  • In a computer system with several possible processing performance states, if the management control function determines that more performance is necessary to meet performance requirements, one approach to providing increased performance is to increase the performance one step at a time until the current utilization is below the high threshold level. However, in another approach rather than increasing the performance state one step at a time, the power management control function selects the highest performance state regardless of the current performance state. The reasons for always selecting the highest possible performance state when a higher performance state is needed are as follows. In computer systems, performance demands are often of a bursty nature. When a higher performance state is required based on the current utilization level, stepping the performance state to a next higher level can result in degradation of performance that can be perceived by the user. That is especially true when the task that needs the increased performance requires a near real-time response, for instance, while decoding an audio or video file. [0022]
  • FIG. 2 illustrates that concept. Assume the processor has five performance states P[0023] 0-P4, with P0 being the highest and P4 the lowest. Whenever the power management determines that a higher performance state is required when operating at any of the levels P1-P4, the power management selects the maximum performance state P0 as the next performance state. Thus, in one embodiment, if the performance state is always taken straight to the maximum performance state when a performance increase is required, rather than stepping up to the maximum performance state, there is less of a chance that a user could notice any performance degradation. In effect, the power management control function anticipates a peak loading by assuming that any indication of a required increase in performance is assumed to be a burst requiring peak performance.
  • However, if a lower performance state is required, a next lower performance state is selected. Thus, if at performance state P[0024] 0, P1 is selected as the next lower performance state. If the current performance state is P1, the next lower performance state selected is P2 when a performance decrease is effectuated by the power management control function. In that way, if the performance is still too high, successively lower performance states can be selected and the chance than any degradation is detected by a system user is reduced. Thus, if the utilization information indicates that an increase in performance is necessary, the power management control function selects the maximum (or near maximum) performance state, while a decrease in performance causes the power management control function to step to the next lowest performance state.
  • If the processor utilization is kept within the range of the high and low thresholds, then a user should experience a crisp, high performance system, while still getting the benefit of power savings for those applications or those portions of applications requiring less performance. That approach reduces power consumption, extends battery life, reduces temperature resulting in less need for cooling and thus less fan noise, while still maintaining high performance and thus maintaining a perception of fast response to the user. Note that running at a lower average CPU die temperature increases CPU reliability, and that a lower CPU temperature results in a lower system temperature, which increases system reliability. [0025]
  • In order to effect changes to the performance state, the power management software has to cause the voltage and frequency used by the CPU to change. In one embodiment that can be accomplished as follows. Referring to FIG. 3, a processor is shown that can dynamically adjust its frequency and/or operating voltage to provide better thermal and power management in accordance with processor utilization. [0026] Processor 301 includes a programmable voltage control field 303, core clock frequency control field 304 and count field 305. Those fields may be located in one or more programmable registers. When the processor and/or system determines that a change to the operating voltage and/or frequency is desired to increase or decrease the performance state, the desired frequency and voltage control information are loaded into frequency control field 304 and voltage control field 303, respectively. Access to a register containing those fields, or an access to another register location, or access to a particular field in one of those registers can be used as a trigger to initiate the protocol for a performance state change. The protocol, which will vary according to processor design, ensures that the processor enters a quiescent state that is suitable for performance state transitions.
  • In some protocols, that will ensure that the portion of the processor, referred to for convenience herein as the [0027] CPU core logic 311 becomes quiescent, including any external bus interfaces 310 and interfaces with internal logic that may not be affected by the performance state change. Note that in one implementation bus interface 310 is implemented separately from CPU core logic 311 and functions to hold off traffic to CPU core logic 311 during performance state transitions. Various messages may be sent or signals provided indicating that the processor has or is about to enter the quiescent state. In addition various control and/or status signals 309 and 313 may be provided to and from core logic 311. Note that in some processor implementations, various power planes and clock grids may ensure some parts of the processor are unaffected by the processor core logic 311 entering the quiescent state. Further, the quiescent state may vary according to processor design. For example, for one processor the quiescent state may have the clocks turned off and in other processors, the clocks may be reduced to an extent to allow appropriate voltage changes. Note also that the power state changes may include clock frequency changes, voltage changes or both. In x86 architectures, such a quiescent state has been referred to as the “stop grant” state in which execution of operating system and application code is stopped. As would be known to those of skill in the art, other ways may be used to initiate the protocol necessary to enter the quiescent state suitable for performance state changes. For example, initiation of the protocol to enter the quiescent state may be indicated by writing a command over a communication link coupling the processor to another device.
  • Once in the quiescent state, the processor can transition the voltage and frequency to the new states specified in [0028] voltage control field 303 and clock frequency control field 304. As described above, in some processor implementations, the processor core clocks are stopped after the processor enters the quiescent state to suitable for changing performance states. In other processor implementations, the processor core clock frequency is reduced to a frequency which can safely tolerate desired voltage changes. In one implementation clock control frequency information is supplied as multiplier values for a clock that is supplied to processor 301. Those of skill in the art appreciate that many other approaches can be used to specify the core operating frequency. In either case, the voltage control information specified in voltage control field 303 is supplied to voltage regulator 315 which in turn supplies CPU core logic 311 with the new voltage during the quiescent state. As will be appreciated by those of skill in the art, there are many alternatives to supplying the voltage information to an external voltage regulator from a register in the processor. For example, the voltage information may be supplied from a device outside the processor, or may be supplied by a communication link with a voltage regulator circuit.
  • Because changing the voltage and frequency can not be done instantaneously, the quiescent state is typically maintained for a period of time to allow the new voltage and clock frequency to stabilize. That time period may be specified in [0029] count field 305 and counted using a counter in the processor or in another location in the computer system. Once the count is complete, the protocol is initiated to exit the quiescent state and resume executing operating system and application code, including restarting any buses that have been “turned off” by the processor as part of entering the quiescent state.
  • Note that changing both voltage and frequency to enter a new performance state can be particularly effective. Changes in the processor's core clock frequency have an approximately linear affect on the power dissipated by the processor. Thus, a 20% reduction in clock frequency reduces the power dissipated by the processor by 20%. The range of change is significant since a ratio of highest frequency to lowest frequency is usually greater than 2:1. Consequently, the processor's power may be changed by similar ratio. Changes in the processor's core voltage have an approximately square law effect. That is, potential power savings is proportional to the square of the voltage reduction. The square law effect results in significant changes in the processor's power if the core voltage of the processor is reduced. [0030]
  • In addition to modifying processor frequency and/or voltage and to match processor utilization, computer systems also monitor temperature and take active or passive power management actions to ensure that the temperature of the processor die stays within operational levels using power management control software such as that associated with the Advanced Configuration and Power Interface (ACPI) Specification. For example, at certain temperatures, the computing system may take actions that take active power management action such as turning on a fan, or increasing fan speed. At other temperatures, a computer system may take passive actions such as throttling the processor (reducing its effective frequency by stopping or slowing clocks with a predetermined duty cycle and period appropriate for the processor). At another temperature threshold, the processor may shut down the computer system to prevent potential damage. FIG. 4 illustrates an exemplary set of temperature “trip points” at which power management activities take place. As shown in FIG. 4, there are “active” trip points at which various actions are performed such as adjusting fan speed. The types of actions described in FIG. 4 are in addition to the utilization based power management described previously. The actual temperatures, the number of trip points, and the specific actions taken will of course vary according to the particulars of the computer system. [0031]
  • Note that the power management control described herein may be implemented as part of the operating system (OS), in whole or in part. In addition, separate power management applications may incorporate some or all of the capability described herein. In other computer systems, the various power management tasks may be shared between the operating system, software applications running under the operating system and BIOS routines. In addition, the power management techniques may be implemented in hardware. [0032]
  • Now that the process of changing processor states has been described, and the utilization of trip points in power management has been described, the utilization of different maximum performance states based on temperature will be described. Assume that a processor system has five performance states P[0033] 0-P4, with P0 being the highest performance state. The highest performance state is that state having the highest clock frequency (and in some embodiments the highest voltage) and having the highest power consumption. Referring again to FIG. 4, the performance states available are shown to be based on temperature range.
  • In the example shown in the last column of the table in FIG. 4, when the die temperature is below 70C, all the performance states are available (P[0034] 0-P4) and the maximum performance state available, e.g., for utilization based (or other) power management activities, is the maximum performance state that the system provides, namely P0. When the temperature is between 70C and 80C, the performance states available are P1-P4 with the maximum available performance state being P1. Finally, when the temperature crosses above 80C the available performance states are further reduced, with the maximum available performance state being P2. Thus, the embodiment represented in FIG. 4, both the number of performance states and the maximum performance state changes based on temperature trip points of 70C and 80C.
  • When the temperature falls back into a temperature range, e.g., <70C, the higher maximum performance state once again becomes available. However, in one embodiment a temperature trip point that causes the system to make a higher maximum performance state available differs from the temperature trip point that causes a lower maximum performance state to be available. That provides hysteresis and avoids frequent changes in the available maximum performance state if the temperature is at or near a trip point. [0035]
  • In the example shown in FIG. 4, hysteresis is implemented by causing P[0036] 0 to become unavailable as the maximum performance state at 70C but to again become available at a temperature trip point of 68C. Thus, if the system is operating with P1 as the maximum performance state and a temperature of 68C is detected, P0 becomes the maximum performance state. Similarly, P1 becomes unavailable as the maximum performance state at 80C but again become available as the maximum performance state when the temperature crosses back below 78C. Note that in various implementations, the system may detect whether the temperature is greater than, greater than or equal to, equal to, less than or less than or equal to a particular temperature, as appropriate for the particular situation, to determine if the system has crossed a temperature trip point.
  • Note that the available performance states may be implemented as a table or other data structure in BIOS, in application software, or in a power management portion of the operating system software. Other ways to encode the appropriate performance states according to the temperature would be readily apparent to those of skill in the art. [0037]
  • The flow chart in FIG. 5 provides a high level illustration of the operation of an embodiment of the invention. For ease of understanding, FIG. 5 deals with maximum performance state transitions and does not illustrate operations of all power management functions. The flow chart corresponds to an embodiment of the invention that utilizes three temperature ranges illustrated in FIG. 4. In [0038] 500, a temperature associated with the microprocessor, e.g., a die temperature of the microprocessor, is supplied or obtained. The system checks to see if it is operating with the maximum performance state being P0 in 503. If so, the system checks to see if the temperature has crossed above the 70C trip point. If so, at 507, the system changes the maximum available performance state to P1 and returns to obtain another temperature reading. If the maximum available performance state is not P0, then the system checks to see at 509 if the maximum performance state is P1. If so, system checks to see if the temperature has crossed below 68C in 511. If so, the maximum performance state is changed to P0 in 513. If not, the system checks in 515 if the temperature has crossed above the 80C trip point. If so, the maximum performance state is reduced to P2 in 517.
  • If the maximum performance state was determined in [0039] 509 to not be P1, then the system determines that the maximum performance state is P2 and the system checks to see if the temperature is less than or equal to the 78C trip point in 521. If so, the maximum performance state is increased to P1 in 523 and the system returns to 500 to obtain a current temperature.
  • Note that the embodiment illustrated in FIG. 5 provides for increasing or decreasing the maximum performance state by one performance state. In other embodiments, the system software and/or hardware may allow for maximum performance state changes of more than one performance state. In addition, while only two trip points are illustrated for reducing a maximum performance state and only two trip points are illustrated for increasing the maximum performance state, the number of trip points can of course vary depending on system design. [0040]
  • In the embodiment illustrated in FIG. 5, hysteresis is implemented at [0041] 511 and 521 such that the temperature at which the system switches back to a higher maximum performance state is different (and lower) than the temperature at which it switched away from that maximum performance state to a lower maximum performance state. While implementation of hysteresis is not necessary to practice the invention, it does prevent frequent switching between maximum allowable performance states when the temperature is hovering around a trip point.
  • If no trip points are crossed, then the maximum performance state remains the same. The process continues periodically evaluating temperatures and setting the appropriate maximum performance state based on the evaluated temperature. In that way, the maximum performance state is determined based on thermal criteria. The system is free to switch to lower performance states according to power management considerations such as processor utilization. [0042]
  • While FIG. 5 provides a high level description of the operation of an embodiment of the invention, determining when temperature thresholds have been crossed may be interrupt driven. For example, temperature sensor apparatus may include programmable high and low thresholds, and generate an interrupt when one of the thresholds is crossed. If the threshold that is crossed causes a maximum performance state change, the high and low thresholds may be set to new threshold values appropriate for the new maximum performance state. The control software is then invoked to change maximum performance states on occurrence of another interrupt. [0043]
  • In comparison to prior art approaches, changing the maximum available performance state according to temperature provides for more gradual reduction in performance as temperature increases. That more gradual reduction in performance results in higher average system performance as temperature increases in comparison to the prior art approach of throttling the clock by a fixed percentage, e.g., 50%, as soon as a first temperature trip point is reached. Thus, a more gradual reduction in performance is provided while still maintaining a high speed rating of the processor in more ideal conditions. [0044]
  • In addition, decreasing the maximum allowable performance state as the temperature goes higher as described herein, allows the case temperature to go higher than solutions that utilize throttling that try to maintain the case temperature at or below a fixed temperature. Allowing the temperature to go higher and not being limited to a fixed case temperature may reduce thermal design costs since the higher allowable case temperature can simplify the thermal design by allowing the system to remove less heat from the system. [0045]
  • As described herein, a computer dynamically adapts its maximum performance state according to a thermal indicator. The description of the invention set forth herein is illustrative, and is not intended to limit the scope of the invention as set forth in the following claims. For instance, while this invention has been described with relation to computer systems such as desktops and a class of mobile computers referred to herein as notebooks (which may also be referred to as laptops or portable computers), the teachings herein may also be utilized in other computing devices, such as servers, multi-processor systems, work stations and/or portable computing devices such as personal digital assistants, (PDAs), which are handheld devices that typically combine computing, telephone/fax, and networking features or in other computing and/or communication equipment, where such power management techniques described herein may prove useful. Other variations and modifications of the embodiments disclosed herein, may be made based on the description set forth herein, without departing from the scope and spirit of the invention as set forth in the following claims. [0046]

Claims (23)

What is claimed is:
1. A method of operating a computing system:
determining a temperature associated with an integrated circuit;
operating the integrated circuit with a first performance state as a maximum performance state according to the determined temperature, the first performance state being one of a plurality of performance states available at the determined temperature.
2. The method as recited in claim 1 wherein the integrated circuit operates at multiple performance states within a temperature range associated with the determined temperature.
3. The method as recited in claim 1 wherein a number of performance states available for operating the integrated circuit varies according to the determined temperature.
4. The method as recited in claim 1 wherein a number of performance states available decreases when the determined temperature crosses above a temperature threshold.
5. The method as recited in claim 1 wherein each performance state is defined by at least one of an operating voltage and frequency of the integrated circuit.
6. The method as recited in claim 1 wherein each performance state is defined at least in part by how much of the integrated circuit is being utilized.
7. The method as recited in claim 1 wherein the integrated circuit is a processor.
8. The method as recited in claim 1 further comprising switching to a second performance state as the maximum performance state when the temperature exceeds a first temperature threshold and wherein the second performance state is a lower performance state than the first performance state.
9. The method as recited in claim 8 further comprising switching back to a higher performance state as the maximum performance state when the temperature is determined to be below a second predetermined temperature threshold below the first temperature threshold.
10. The method as recited in claim 9 wherein the higher performance sate is the first performance state.
11. The method as recited in claim 1 further comprising switching to a second performance state as the maximum performance state when the temperature is determined to be below a predetermined temperature threshold and wherein the second performance state is a higher performance state than the first performance state.
12. The method as recited in claim 8 wherein the determination that the temperature exceeds a first temperature threshold is a determination that the temperature has moved from a first temperature range associated with the first performance state being the maximum performance state to a second temperature range associated with the second performance state being the maximum operating state.
13. The method as recited in claim 12 wherein the first and second temperature ranges have a boundary at the first temperature threshold.
14. A computing system comprising:
an integrated circuit operable at multiple performance states, the performance states being defined by at least one of operating voltage and frequency;
and wherein the computing system provides that the integrated circuit, at a first detected temperature, has a first maximum performance state and a first plurality of lesser performance states; and wherein at a second detected temperature, higher than the first detected temperature, the integrated circuit has a lower maximum performance state and a second plurality of lesser performance states, the lower maximum performance state providing lower performance than the first maximum performance state in terms of maximum power consumption.
15. The computing system as recited in claim 14 further comprising:
a temperature detection mechanism coupled to detect a temperature associated with the integrated circuit; and
wherein the computing system is operable to change to a different maximum performance state according to the detected temperature.
16. The computing system as recited in claim 15 wherein the detected temperature is one of ambient temperature, junction temperature, case temperature, or die temperature.
17. The computing system as recited in claim 15 wherein a higher detected temperature results in a lower maximum performance state.
18. The computing system as recited in claim 15 wherein the number of performance states available varies according to the detected temperature.
19. The computing system as recited in claim 15 wherein the integrated circuit is a processor.
20. A computer program product encoded in at least one computer readable medium, the computer program product comprising:
a plurality of groups of performance operating states, each of the groups of performance operating states having a different maximum operating state, the groups of operating states corresponding to different temperature related to a processor; and
an instruction sequence executable to change to a different group of performance operating states and thereby a different maximum operating state according to a detected temperature associated with the computer system.
21. The computer program product of 20, wherein the at least one computer readable medium is selected from the set of a disk, tape or other magnetic, optical, or electronic storage medium and a network, wire line, wireless or other communications medium.
22. A computing system comprising:
means for determining a temperature associated with a processor, the processor having a plurality of groups of performance states associated with each of a plurality of temperature ranges, each of the groups having a different maximum performance state; and
means for changing from a first group of performance states to a second group of performance states according to the determined temperature, thereby changing the available maximum performance state available for processor operation
23. A method of providing a variable maximum die temperature for an integrated circuit in a computer system having a plurality of different maximum performance states, comprising increasing the maximum allowable die temperature as a maximum allowable performance state is decreased.
US10/015,031 2001-12-11 2001-12-11 Variable maximum die temperature based on performance state Expired - Lifetime US6889332B2 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US10/015,031 US6889332B2 (en) 2001-12-11 2001-12-11 Variable maximum die temperature based on performance state

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
US10/015,031 US6889332B2 (en) 2001-12-11 2001-12-11 Variable maximum die temperature based on performance state

Publications (2)

Publication Number Publication Date
US20030110423A1 true US20030110423A1 (en) 2003-06-12
US6889332B2 US6889332B2 (en) 2005-05-03

Family

ID=21769165

Family Applications (1)

Application Number Title Priority Date Filing Date
US10/015,031 Expired - Lifetime US6889332B2 (en) 2001-12-11 2001-12-11 Variable maximum die temperature based on performance state

Country Status (1)

Country Link
US (1) US6889332B2 (en)

Cited By (78)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20020087896A1 (en) * 2000-12-29 2002-07-04 Cline Leslie E. Processor performance state control
US20030126479A1 (en) * 2001-12-28 2003-07-03 Burns James S. Digital throttle for multiple operating points
US20030140265A1 (en) * 2001-12-20 2003-07-24 Michael Henninger Method for controlling a safety-critical system which has a microcontroller
US20040268174A1 (en) * 2001-12-12 2004-12-30 Barnes Cooper Operating system coordinated thermal management
US20050071705A1 (en) * 2003-09-29 2005-03-31 Ati Technologies, Inc. Adaptive temperature dependent feedback clock control system and method
US20060052970A1 (en) * 2004-08-23 2006-03-09 Tawfik Arabi On-die temperature control data for communicating to a thermal actuator
US20060075265A1 (en) * 2004-09-22 2006-04-06 Yoshiyuki Hamaoka Self-organized parallel processing system
US20060100798A1 (en) * 2004-11-05 2006-05-11 Schweitzer Engineering Laboratories, Inc. Method to increase the maximum allowable ambient temperature rating of an electronic device
US20060149974A1 (en) * 2004-12-30 2006-07-06 Efraim Rotem Device and method for on-die temperature measurement
GB2422461A (en) * 2005-01-21 2006-07-26 Hewlett Packard Development Co Maintaining the performance of an integrated circuit
US20070005152A1 (en) * 2005-06-30 2007-01-04 Ben Karr Method and apparatus for monitoring power in integrated circuits
US20070058547A1 (en) * 2005-09-13 2007-03-15 Viktors Berstis Method and apparatus for a grid network throttle and load collector
US20070250219A1 (en) * 2002-10-03 2007-10-25 Via Technologies, Inc. Microprocessor capable of dynamically reducing its power consumption in response to varying operating temperature
US20070250736A1 (en) * 2002-10-03 2007-10-25 Via Technologies, Inc. Microprocessor with improved performance during p-state transitions
US20070255972A1 (en) * 2002-10-03 2007-11-01 Via Technologies, Inc. Microprocessor capable of dynamically increasing its performance in response to varying operating temperature
US20080011467A1 (en) * 2006-06-23 2008-01-17 Intel Corporation Method, apparatus and system for thermal management using power density feedback
US20080040622A1 (en) * 2006-08-14 2008-02-14 Advanced Micro Devices, Inc. System and method for limiting processor performance
US20080036613A1 (en) * 2002-10-03 2008-02-14 Via Technologies, Inc. Microprocessor with improved thermal monitoring and protection mechanism
US20080141047A1 (en) * 2004-09-10 2008-06-12 Freescale Semiconductor, Inc. Apparatus and Method for Controlling Voltage and Frequency
CN100422942C (en) * 2004-11-04 2008-10-01 株式会社东芝 Processor system with temperature sensor and control method of the same
US7479753B1 (en) 2004-02-24 2009-01-20 Nvidia Corporation Fan speed controller
US20090144572A1 (en) * 2004-09-10 2009-06-04 Freescale Semiconductor, Inc. Apparatus and method for controlling voltage and frequency
US7793291B2 (en) 2004-12-22 2010-09-07 International Business Machines Corporation Thermal management of a multi-processor computer system
US7882369B1 (en) 2002-11-14 2011-02-01 Nvidia Corporation Processor performance adjustment system and method
US7886164B1 (en) 2002-11-14 2011-02-08 Nvidia Corporation Processor temperature adjustment system and method
US20120096288A1 (en) * 2010-10-13 2012-04-19 Bates Benjamin D Controlling operation of temperature sensors
US8370663B2 (en) 2008-02-11 2013-02-05 Nvidia Corporation Power management with dynamic frequency adjustments
US20130086395A1 (en) * 2011-09-30 2013-04-04 Qualcomm Incorporated Multi-Core Microprocessor Reliability Optimization
US20130178999A1 (en) * 2012-01-09 2013-07-11 International Business Machines Corporation Managing workload distribution among computing systems to optimize heat dissipation by computing systems
US20140201562A1 (en) * 2010-10-10 2014-07-17 Liqid Inc. Systems and methods for optimizing data storage among a plurality of solid state memory subsystems
US20140222242A1 (en) * 2011-12-29 2014-08-07 Rajesh Poornachandran Adaptive thermal throttling with user configuration capability
US8839006B2 (en) 2010-05-28 2014-09-16 Nvidia Corporation Power consumption reduction systems and methods
CN104049714A (en) * 2013-03-15 2014-09-17 英特尔公司 Processor having frequency of operation information for guaranteed operation under high temperature events
US8972754B2 (en) 2011-09-08 2015-03-03 Asustek Computer Inc. Computer device and frequency adjusting method for central processing unit
US8996902B2 (en) 2012-10-23 2015-03-31 Qualcomm Incorporated Modal workload scheduling in a heterogeneous multi-processor system on a chip
EP1965286A3 (en) * 2007-03-01 2015-04-01 VIA Technologies, Inc. Microprocessor with improved thermal monitoring and protection mechanism
US20150098476A1 (en) * 2013-10-03 2015-04-09 International Business Machines Corporation Temperature sensitive routing of data in a computer system
US20150169016A1 (en) * 2012-06-29 2015-06-18 Qualcomm Incorporated System and method for adaptive thermal management in a portable computing device
US9134782B2 (en) 2007-05-07 2015-09-15 Nvidia Corporation Maintaining optimum voltage supply to match performance of an integrated circuit
US9158311B2 (en) 2011-12-14 2015-10-13 International Business Machines Corporation Integrating a data center thermal control system and individual fan controllers for controlling a thermal environment in a data center room
US20150331433A1 (en) * 2014-05-14 2015-11-19 Advanced Micro Devices, Inc. Hybrid system and method for determining performance levels based on thermal conditions within a processor
US9256265B2 (en) 2009-12-30 2016-02-09 Nvidia Corporation Method and system for artificially and dynamically limiting the framerate of a graphics processing unit
CN105334939A (en) * 2014-06-30 2016-02-17 北京安兔兔科技有限公司 Terminal power consumption control method and terminal power consumption control device
US20160274629A1 (en) * 2015-03-20 2016-09-22 Dominick Adam Lovicott Systems And Methods Of Adaptive Thermal Control For Information Handling Systems
EP2796961A3 (en) * 2013-04-25 2016-10-26 Intel Corporation Controlling power and performance in a system agent of a processor
US20170017608A1 (en) * 2015-07-16 2017-01-19 Motorola Mobility Llc Apparatus and method for improving application performance within a group of communication devices
US9645902B2 (en) 2014-06-23 2017-05-09 Liqid Inc. Modular switched fabric for data storage systems
US9652019B2 (en) 2014-06-02 2017-05-16 Advanced Micro Devices, Inc. System and method for adjusting processor performance based on platform and ambient thermal conditions
US9678910B2 (en) 2014-04-25 2017-06-13 Liqid Inc. Power handling in a scalable storage system
CN107153592A (en) * 2016-03-02 2017-09-12 神讯电脑(昆山)有限公司 Electronic installation and its power management method
US9830889B2 (en) 2009-12-31 2017-11-28 Nvidia Corporation Methods and system for artifically and dynamically limiting the display resolution of an application
EP3291048A1 (en) * 2016-09-05 2018-03-07 Intel IP Corporation Method and device for thermal management control of an electronic device
US10001819B2 (en) 2014-09-04 2018-06-19 Liqid Inc. Dual-sided rackmount modular data processing assembly
US10001800B1 (en) 2015-09-10 2018-06-19 Apple Inc. Systems and methods for determining temperatures of integrated circuits
US10019388B2 (en) 2015-04-28 2018-07-10 Liqid Inc. Enhanced initialization for data storage assemblies
US20180203494A1 (en) * 2015-07-29 2018-07-19 Zte Corporation Method and apparatus for controlling hot plug operation of cpu in mobile terminal
US10108422B2 (en) 2015-04-28 2018-10-23 Liqid Inc. Multi-thread network stack buffering of data frames
US10175748B2 (en) * 2017-01-05 2019-01-08 Nuvoton Technology Corporation System, method and computer program product for improved regulation of an electrical device's consumption of power from an external power supply
US10180924B2 (en) 2017-05-08 2019-01-15 Liqid Inc. Peer-to-peer communication for graphics processing units
US10191691B2 (en) 2015-04-28 2019-01-29 Liqid Inc. Front-end quality of service differentiation in storage system operations
US20190034241A1 (en) * 2016-08-16 2019-01-31 International Business Machines Corporation Optimized resource metering in a multi tenanted distributed file system
US10198183B2 (en) 2015-02-06 2019-02-05 Liqid Inc. Tunneling of storage operations between storage nodes
US10255215B2 (en) 2016-01-29 2019-04-09 Liqid Inc. Enhanced PCIe storage device form factors
US10362107B2 (en) 2014-09-04 2019-07-23 Liqid Inc. Synchronization of storage transactions in clustered storage systems
US10402363B2 (en) 2016-06-10 2019-09-03 Liqid Inc. Multi-port interposer architectures in data storage systems
US10467166B2 (en) 2014-04-25 2019-11-05 Liqid Inc. Stacked-device peripheral storage card
US10585827B1 (en) 2019-02-05 2020-03-10 Liqid Inc. PCIe fabric enabled peer-to-peer communications
US10592291B2 (en) 2016-08-12 2020-03-17 Liqid Inc. Disaggregated fabric-switched computing platform
US10614022B2 (en) 2017-04-27 2020-04-07 Liqid Inc. PCIe fabric connectivity expansion card
US10660228B2 (en) 2018-08-03 2020-05-19 Liqid Inc. Peripheral storage card with offset slot alignment
US11086371B2 (en) * 2018-06-06 2021-08-10 Lenovo (Singapore) Pte. Ltd. Heat suppression in an information processing device in an initial setting mode
US11256649B2 (en) 2019-04-25 2022-02-22 Liqid Inc. Machine templates for predetermined compute units
US11265219B2 (en) 2019-04-25 2022-03-01 Liqid Inc. Composed computing systems with converged and disaggregated component pool
US11294839B2 (en) 2016-08-12 2022-04-05 Liqid Inc. Emulated telemetry interfaces for fabric-coupled computing units
US11399720B2 (en) 2016-04-05 2022-08-02 Qulacomm Incorporated Circuits and methods providing temperature mitigation for computing devices
US11442776B2 (en) 2020-12-11 2022-09-13 Liqid Inc. Execution job compute unit composition in computing clusters
US11880326B2 (en) 2016-08-12 2024-01-23 Liqid Inc. Emulated telemetry interfaces for computing units
WO2024183598A1 (en) * 2023-03-03 2024-09-12 华为技术有限公司 Frequency control method for processor, and electronic device

Families Citing this family (38)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6895520B1 (en) 2001-03-02 2005-05-17 Advanced Micro Devices, Inc. Performance and power optimization via block oriented performance measurement and control
US6893979B2 (en) * 2001-03-15 2005-05-17 International Business Machines Corporation Method for improved plasma nitridation of ultra thin gate dielectrics
US7231531B2 (en) * 2001-03-16 2007-06-12 Dualcor Technologies, Inc. Personal electronics device with a dual core processor
US7020786B2 (en) * 2002-07-23 2006-03-28 Dell Products L.P. System and method for selecting a voltage output reference
KR20050085962A (en) * 2003-01-13 2005-08-29 에이알엠 리미티드 Data processing performance control
US7085945B2 (en) * 2003-01-24 2006-08-01 Intel Corporation Using multiple thermal points to enable component level power and thermal management
DE602004007525T2 (en) * 2003-07-09 2008-03-13 Nxp B.V. IC WITH INTEGRATED CHARACTERISTIC PROVISION
US8237386B2 (en) 2003-08-15 2012-08-07 Apple Inc. Methods and apparatuses for operating a data processing system
US7308590B2 (en) * 2004-10-15 2007-12-11 Intel Corporation Automatic dynamic processor operating voltage control
US20060161375A1 (en) * 2004-12-30 2006-07-20 Allen Duberstein Optimizing processing speed based on measured temperatures
US7194645B2 (en) * 2005-02-09 2007-03-20 International Business Machines Corporation Method and apparatus for autonomic policy-based thermal management in a data processing system
US7562234B2 (en) 2005-08-25 2009-07-14 Apple Inc. Methods and apparatuses for dynamic power control
US7337339B1 (en) 2005-09-15 2008-02-26 Azul Systems, Inc. Multi-level power monitoring, filtering and throttling at local blocks and globally
US7577860B2 (en) * 2006-01-09 2009-08-18 Microsoft Corporation Processor specific BIOS interface for power management
US7886167B2 (en) * 2006-05-11 2011-02-08 Intel Corporation Load circuit supply voltage control
US8044697B2 (en) * 2006-06-29 2011-10-25 Intel Corporation Per die temperature programming for thermally efficient integrated circuit (IC) operation
US20090049459A1 (en) * 2007-08-14 2009-02-19 Microsoft Corporation Dynamically converting symbolic links
US8078890B2 (en) * 2007-09-11 2011-12-13 Dell Products L.L.P. System and method for providing memory performance states in a computing system
US8001402B2 (en) * 2007-12-05 2011-08-16 International Business Machines Corporation Method for power capping with co-operative dynamic voltage and frequency scaling via shared p-state table
US8341433B2 (en) * 2008-01-04 2012-12-25 Dell Products L.P. Method and system for managing the power consumption of an information handling system
US20090235108A1 (en) * 2008-03-11 2009-09-17 Gold Spencer M Automatic processor overclocking
US20090256622A1 (en) * 2008-04-11 2009-10-15 Nortel Networks Limited Soft thermal failure in a high capacity transmission system
US8037325B1 (en) 2008-12-09 2011-10-11 Google Inc. System and method for determining power consumption
US8458498B2 (en) * 2008-12-23 2013-06-04 Intel Corporation Method and apparatus of power management of processor
US8064197B2 (en) * 2009-05-22 2011-11-22 Advanced Micro Devices, Inc. Heat management using power management information
US8301873B2 (en) * 2009-12-01 2012-10-30 Getac Technology Corporation Method and computer system for thermal throttling protection
US8909961B2 (en) 2011-11-29 2014-12-09 Ati Technologies Ulc Method and apparatus for adjusting power consumption level of an integrated circuit
US9223383B2 (en) 2012-12-21 2015-12-29 Advanced Micro Devices, Inc. Guardband reduction for multi-core data processor
US9360918B2 (en) 2012-12-21 2016-06-07 Advanced Micro Devices, Inc. Power control for multi-core data processor
US9483092B2 (en) 2013-10-14 2016-11-01 Advanced Micro Devices, Inc. Performance state boost for multi-core integrated circuit
US9625986B2 (en) * 2015-03-13 2017-04-18 Kabushiki Kaisha Toshiba Semiconductor device and temperature control method of semiconductor device
US10248186B2 (en) 2016-06-10 2019-04-02 Microsoft Technology Licensing, Llc Processor device voltage characterization
US10209726B2 (en) 2016-06-10 2019-02-19 Microsoft Technology Licensing, Llc Secure input voltage adjustment in processing devices
US10310572B2 (en) 2016-06-10 2019-06-04 Microsoft Technology Licensing, Llc Voltage based thermal control of processing device
US10338670B2 (en) 2016-06-10 2019-07-02 Microsoft Technology Licensing, Llc Input voltage reduction for processing devices
US10572183B2 (en) * 2017-10-18 2020-02-25 Advanced Micro Devices, Inc. Power efficient retraining of memory accesses
US11892893B2 (en) 2019-10-01 2024-02-06 Microsoft Technology Licensing, Llc Systems and methods for thermal system management
US11720158B2 (en) 2020-03-13 2023-08-08 Google Llc Power throttling mechanism using instruction rate limiting in high power machine-learning ASICs

Citations (21)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US36839A (en) * 1862-11-04 gordan
US5502838A (en) * 1994-04-28 1996-03-26 Consilium Overseas Limited Temperature management for integrated circuits
US5511203A (en) * 1994-02-02 1996-04-23 Advanced Micro Devices Power management system distinguishing between primary and secondary system activity
US5682273A (en) * 1995-06-30 1997-10-28 International Business Machines Corporation Disk drive for portable computer with adaptive demand-driven power management
US5713030A (en) * 1995-10-11 1998-01-27 Vlsi Technology, Inc. Thermal management device and method for a computer processor
US5745375A (en) * 1995-09-29 1998-04-28 Intel Corporation Apparatus and method for controlling power usage
US5752011A (en) * 1994-06-20 1998-05-12 Thomas; C. Douglas Method and system for controlling a processor's clock frequency in accordance with the processor's temperature
US5812860A (en) * 1996-02-12 1998-09-22 Intel Corporation Method and apparatus providing multiple voltages and frequencies selectable based on real time criteria to control power consumption
US5818968A (en) * 1995-03-20 1998-10-06 Sony Corporation High-efficiency coding method, high-efficiency coding apparatus, recording and reproducing apparatus, and information transmission system
US5852737A (en) * 1995-04-24 1998-12-22 National Semiconductor Corporation Method and apparatus for operating digital static CMOS components in a very low voltage mode during power-down
US5873000A (en) * 1996-07-19 1999-02-16 Compaq Computer Corporation System incorporating hot docking and undocking capabilities without requiring a standby or suspend mode by placing local arbiters of system and base into idle state
US5881298A (en) * 1996-09-05 1999-03-09 Micron Technology, Inc. Portable computer with selectively operable cooling unit
US5884049A (en) * 1996-12-31 1999-03-16 Compaq Computer Corporation Increased processor performance comparable to a desktop computer from a docked portable computer
US5887179A (en) * 1996-06-11 1999-03-23 Motorola, Inc. System power saving means and method
US5925133A (en) * 1994-10-19 1999-07-20 Advanced Micro Devices, Inc. Integrated processor system adapted for portable personal information devices
US5954820A (en) * 1997-09-26 1999-09-21 International Business Machines Corporation Portable computer with adaptive demand-driven power management
US6119241A (en) * 1996-12-23 2000-09-12 International Business Machines Corporation Self regulating temperature/performance/voltage scheme for micros (X86)
US6363490B1 (en) * 1999-03-30 2002-03-26 Intel Corporation Method and apparatus for monitoring the temperature of a processor
US6630754B1 (en) * 1993-09-21 2003-10-07 Intel Corporation Temperature-based cooling device controller apparatus and method
US6647320B1 (en) * 1999-09-13 2003-11-11 Fujitsu Limited Software-based temperature controller circuit in electronic apparatus
US6701273B2 (en) * 1994-03-28 2004-03-02 Kabushiki Kaisha Toshiba Method and apparatus for controlling internal heat generating circuitry

Family Cites Families (14)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0632360A1 (en) 1993-06-29 1995-01-04 Xerox Corporation Reducing computer power consumption by dynamic voltage and frequency variation
WO1996025701A1 (en) 1995-02-14 1996-08-22 Vlsi Technology, Inc. Method and apparatus for reducing power consumption in digital electronic circuits
JP3075957B2 (en) 1995-05-30 2000-08-14 株式会社東芝 Computer system
US5838968A (en) 1996-03-01 1998-11-17 Chromatic Research, Inc. System and method for dynamic resource management across tasks in real-time operating systems
US6006248A (en) 1996-07-12 1999-12-21 Nec Corporation Job application distributing system among a plurality of computers, job application distributing method and recording media in which job application distributing program is recorded
JPH10268963A (en) 1997-03-28 1998-10-09 Mitsubishi Electric Corp Information processor
US6151681A (en) 1997-06-25 2000-11-21 Texas Instruments Incorporated Dynamic device power management
US6035388A (en) 1997-06-27 2000-03-07 Sandcraft, Inc. Method and apparatus for dual issue of program instructions to symmetric multifunctional execution units
US5958058A (en) 1997-07-18 1999-09-28 Micron Electronics, Inc. User-selectable power management interface with application threshold warnings
JPH11184554A (en) 1997-12-24 1999-07-09 Mitsubishi Electric Corp Clock control type information processor
US6128745A (en) 1998-05-28 2000-10-03 Phoenix Technologies Ltd. Power management inactivity monitoring using software threads
JP3297389B2 (en) 1998-12-07 2002-07-02 インターナショナル・ビジネス・マシーンズ・コーポレーション Power consumption control method and electric equipment
US6442700B1 (en) 1999-08-10 2002-08-27 Intel Corporation Thermal control within systems having multiple CPU performance states
US6711129B1 (en) 1999-10-26 2004-03-23 Avaya Technology Corp. Real-time admission control

Patent Citations (21)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US36839A (en) * 1862-11-04 gordan
US6630754B1 (en) * 1993-09-21 2003-10-07 Intel Corporation Temperature-based cooling device controller apparatus and method
US5511203A (en) * 1994-02-02 1996-04-23 Advanced Micro Devices Power management system distinguishing between primary and secondary system activity
US6701273B2 (en) * 1994-03-28 2004-03-02 Kabushiki Kaisha Toshiba Method and apparatus for controlling internal heat generating circuitry
US5502838A (en) * 1994-04-28 1996-03-26 Consilium Overseas Limited Temperature management for integrated circuits
US5752011A (en) * 1994-06-20 1998-05-12 Thomas; C. Douglas Method and system for controlling a processor's clock frequency in accordance with the processor's temperature
US5925133A (en) * 1994-10-19 1999-07-20 Advanced Micro Devices, Inc. Integrated processor system adapted for portable personal information devices
US5818968A (en) * 1995-03-20 1998-10-06 Sony Corporation High-efficiency coding method, high-efficiency coding apparatus, recording and reproducing apparatus, and information transmission system
US5852737A (en) * 1995-04-24 1998-12-22 National Semiconductor Corporation Method and apparatus for operating digital static CMOS components in a very low voltage mode during power-down
US5682273A (en) * 1995-06-30 1997-10-28 International Business Machines Corporation Disk drive for portable computer with adaptive demand-driven power management
US5745375A (en) * 1995-09-29 1998-04-28 Intel Corporation Apparatus and method for controlling power usage
US5713030A (en) * 1995-10-11 1998-01-27 Vlsi Technology, Inc. Thermal management device and method for a computer processor
US5812860A (en) * 1996-02-12 1998-09-22 Intel Corporation Method and apparatus providing multiple voltages and frequencies selectable based on real time criteria to control power consumption
US5887179A (en) * 1996-06-11 1999-03-23 Motorola, Inc. System power saving means and method
US5873000A (en) * 1996-07-19 1999-02-16 Compaq Computer Corporation System incorporating hot docking and undocking capabilities without requiring a standby or suspend mode by placing local arbiters of system and base into idle state
US5881298A (en) * 1996-09-05 1999-03-09 Micron Technology, Inc. Portable computer with selectively operable cooling unit
US6119241A (en) * 1996-12-23 2000-09-12 International Business Machines Corporation Self regulating temperature/performance/voltage scheme for micros (X86)
US5884049A (en) * 1996-12-31 1999-03-16 Compaq Computer Corporation Increased processor performance comparable to a desktop computer from a docked portable computer
US5954820A (en) * 1997-09-26 1999-09-21 International Business Machines Corporation Portable computer with adaptive demand-driven power management
US6363490B1 (en) * 1999-03-30 2002-03-26 Intel Corporation Method and apparatus for monitoring the temperature of a processor
US6647320B1 (en) * 1999-09-13 2003-11-11 Fujitsu Limited Software-based temperature controller circuit in electronic apparatus

Cited By (171)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20020087896A1 (en) * 2000-12-29 2002-07-04 Cline Leslie E. Processor performance state control
US6988211B2 (en) * 2000-12-29 2006-01-17 Intel Corporation System and method for selecting a frequency and voltage combination from a table using a selection field and a read-only limit field
US7054720B2 (en) * 2001-12-12 2006-05-30 Intel Corporation Operating system coordinated thermal management
US20040268174A1 (en) * 2001-12-12 2004-12-30 Barnes Cooper Operating system coordinated thermal management
US20030140265A1 (en) * 2001-12-20 2003-07-24 Michael Henninger Method for controlling a safety-critical system which has a microcontroller
US7171582B2 (en) * 2001-12-20 2007-01-30 Siemens Aktiengesellschaft Method for controlling a safety-critical system which has a microcontroller
US20030126479A1 (en) * 2001-12-28 2003-07-03 Burns James S. Digital throttle for multiple operating points
US7281140B2 (en) * 2001-12-28 2007-10-09 Intel Corporation Digital throttle for multiple operating points
US20100324750A1 (en) * 2002-10-03 2010-12-23 Via Technologies, Inc. Microprocessor with improved thermal monitoring and protection mechanism
US7770042B2 (en) 2002-10-03 2010-08-03 Via Technologies, Inc. Microprocessor with improved performance during P-state transitions
US20080036613A1 (en) * 2002-10-03 2008-02-14 Via Technologies, Inc. Microprocessor with improved thermal monitoring and protection mechanism
US20070255972A1 (en) * 2002-10-03 2007-11-01 Via Technologies, Inc. Microprocessor capable of dynamically increasing its performance in response to varying operating temperature
US7814350B2 (en) * 2002-10-03 2010-10-12 Via Technologies, Inc. Microprocessor with improved thermal monitoring and protection mechanism
US20070250736A1 (en) * 2002-10-03 2007-10-25 Via Technologies, Inc. Microprocessor with improved performance during p-state transitions
US7774627B2 (en) * 2002-10-03 2010-08-10 Via Technologies, Inc. Microprocessor capable of dynamically increasing its performance in response to varying operating temperature
US8412962B2 (en) 2002-10-03 2013-04-02 Via Technologies, Inc. Microprocessor with improved thermal monitoring and protection mechanism
US7698583B2 (en) 2002-10-03 2010-04-13 Via Technologies, Inc. Microprocessor capable of dynamically reducing its power consumption in response to varying operating temperature
US20070250219A1 (en) * 2002-10-03 2007-10-25 Via Technologies, Inc. Microprocessor capable of dynamically reducing its power consumption in response to varying operating temperature
US7886164B1 (en) 2002-11-14 2011-02-08 Nvidia Corporation Processor temperature adjustment system and method
US7882369B1 (en) 2002-11-14 2011-02-01 Nvidia Corporation Processor performance adjustment system and method
EP1519259A3 (en) * 2003-09-29 2006-08-16 ATI Technologies Inc. Adaptive temperature dependent feedback clock control system and method
US20050071705A1 (en) * 2003-09-29 2005-03-31 Ati Technologies, Inc. Adaptive temperature dependent feedback clock control system and method
US7467318B2 (en) 2003-09-29 2008-12-16 Ati Technologies Ulc Adaptive temperature dependent feedback clock control system and method
US7479753B1 (en) 2004-02-24 2009-01-20 Nvidia Corporation Fan speed controller
US20060052970A1 (en) * 2004-08-23 2006-03-09 Tawfik Arabi On-die temperature control data for communicating to a thermal actuator
US7117114B2 (en) * 2004-08-23 2006-10-03 Intel Corporation On-die temperature control data for communicating to a thermal actuator
US20080141047A1 (en) * 2004-09-10 2008-06-12 Freescale Semiconductor, Inc. Apparatus and Method for Controlling Voltage and Frequency
US20090144572A1 (en) * 2004-09-10 2009-06-04 Freescale Semiconductor, Inc. Apparatus and method for controlling voltage and frequency
US7975155B2 (en) 2004-09-10 2011-07-05 Freescale Semiconductor, Inc. Apparatus and method for controlling voltage and frequency
US20060075265A1 (en) * 2004-09-22 2006-04-06 Yoshiyuki Hamaoka Self-organized parallel processing system
US7543091B2 (en) * 2004-09-22 2009-06-02 Kabushiki Kaisha Toshiba Self-organized parallel processing system
US7913006B2 (en) 2004-09-22 2011-03-22 Kabushiki Kaisha Toshiba Self-organized parallel processing system
US20090172016A1 (en) * 2004-09-22 2009-07-02 Yoshiyuki Hamaoka Self-Organized Parallel Processing System
CN100422942C (en) * 2004-11-04 2008-10-01 株式会社东芝 Processor system with temperature sensor and control method of the same
US20060100798A1 (en) * 2004-11-05 2006-05-11 Schweitzer Engineering Laboratories, Inc. Method to increase the maximum allowable ambient temperature rating of an electronic device
US7313500B2 (en) * 2004-11-05 2007-12-25 Schweitzer Engineering Labortories, Inc. Method to increase the maximum allowable ambient temperature rating of an electronic device
US7793291B2 (en) 2004-12-22 2010-09-07 International Business Machines Corporation Thermal management of a multi-processor computer system
US20060149974A1 (en) * 2004-12-30 2006-07-06 Efraim Rotem Device and method for on-die temperature measurement
US7878016B2 (en) * 2004-12-30 2011-02-01 Intel Corporation Device and method for on-die temperature measurement
US7661003B2 (en) * 2005-01-21 2010-02-09 Hewlett-Packard Development Company, L.P. Systems and methods for maintaining performance of an integrated circuit within a working power limit
US20060167657A1 (en) * 2005-01-21 2006-07-27 Naffziger Samuel D Systems and methods for maintaining performance
GB2422461B (en) * 2005-01-21 2008-09-03 Hewlett Packard Development Co Systems and methods for maintaining performance
GB2422461A (en) * 2005-01-21 2006-07-26 Hewlett Packard Development Co Maintaining the performance of an integrated circuit
US20070005152A1 (en) * 2005-06-30 2007-01-04 Ben Karr Method and apparatus for monitoring power in integrated circuits
US20070058547A1 (en) * 2005-09-13 2007-03-15 Viktors Berstis Method and apparatus for a grid network throttle and load collector
US7995474B2 (en) * 2005-09-13 2011-08-09 International Business Machines Corporation Grid network throttle and load collector
US20080011467A1 (en) * 2006-06-23 2008-01-17 Intel Corporation Method, apparatus and system for thermal management using power density feedback
TWI411913B (en) * 2006-08-14 2013-10-11 Globalfoundries Us Inc System and method for limiting processor performance
US7747881B2 (en) * 2006-08-14 2010-06-29 Globalfoundries Inc. System and method for limiting processor performance
US20080040622A1 (en) * 2006-08-14 2008-02-14 Advanced Micro Devices, Inc. System and method for limiting processor performance
GB2454400B (en) * 2006-08-14 2011-02-23 Advanced Micro Devices Inc System and method for limiting processor performance
EP1965286A3 (en) * 2007-03-01 2015-04-01 VIA Technologies, Inc. Microprocessor with improved thermal monitoring and protection mechanism
EP1965287A3 (en) * 2007-03-02 2010-11-24 VIA Technologies, Inc. Microprocessor capable of dynamically increasing its performance in response to varying operating temperature
EP1965287A2 (en) * 2007-03-02 2008-09-03 VIA Technologies, Inc. Microprocessor capable of dynamically increasing its performance in response to varying operating temperature
US9134782B2 (en) 2007-05-07 2015-09-15 Nvidia Corporation Maintaining optimum voltage supply to match performance of an integrated circuit
US8370663B2 (en) 2008-02-11 2013-02-05 Nvidia Corporation Power management with dynamic frequency adjustments
US8775843B2 (en) 2008-02-11 2014-07-08 Nvidia Corporation Power management with dynamic frequency adjustments
US9256265B2 (en) 2009-12-30 2016-02-09 Nvidia Corporation Method and system for artificially and dynamically limiting the framerate of a graphics processing unit
US9830889B2 (en) 2009-12-31 2017-11-28 Nvidia Corporation Methods and system for artifically and dynamically limiting the display resolution of an application
US8839006B2 (en) 2010-05-28 2014-09-16 Nvidia Corporation Power consumption reduction systems and methods
US20140201562A1 (en) * 2010-10-10 2014-07-17 Liqid Inc. Systems and methods for optimizing data storage among a plurality of solid state memory subsystems
US10191667B2 (en) 2010-10-10 2019-01-29 Liqid Inc. Systems and methods for optimizing data storage among a plurality of storage drives
US9285827B2 (en) * 2010-10-10 2016-03-15 Liqid Inc. Systems and methods for optimizing data storage among a plurality of solid state memory subsystems
US10795584B2 (en) 2010-10-10 2020-10-06 Liqid Inc. Data storage among a plurality of storage drives
US11366591B2 (en) 2010-10-10 2022-06-21 Liqid Inc. Data storage among a plurality of storage drives
US9317082B2 (en) * 2010-10-13 2016-04-19 Advanced Micro Devices, Inc. Controlling operation of temperature sensors
US20120096288A1 (en) * 2010-10-13 2012-04-19 Bates Benjamin D Controlling operation of temperature sensors
US8972754B2 (en) 2011-09-08 2015-03-03 Asustek Computer Inc. Computer device and frequency adjusting method for central processing unit
US20130086395A1 (en) * 2011-09-30 2013-04-04 Qualcomm Incorporated Multi-Core Microprocessor Reliability Optimization
US9158310B2 (en) 2011-12-14 2015-10-13 International Business Machines Corporation Integrating a data center thermal control system and individual fan controllers for controlling a thermal environment in a data center room
US9158311B2 (en) 2011-12-14 2015-10-13 International Business Machines Corporation Integrating a data center thermal control system and individual fan controllers for controlling a thermal environment in a data center room
US20140222242A1 (en) * 2011-12-29 2014-08-07 Rajesh Poornachandran Adaptive thermal throttling with user configuration capability
US9798335B2 (en) * 2011-12-29 2017-10-24 Intel Corporation Adaptive thermal throttling with user configuration capability
US20140052311A1 (en) * 2012-01-09 2014-02-20 International Business Machines Corporation Managing workload distribution among computing systems to optimize heat dissipation by computing systems
US9176508B2 (en) * 2012-01-09 2015-11-03 International Business Machines Corporation Managing workload distribution among computing systems to optimize heat dissipation by computing systems
US20130178999A1 (en) * 2012-01-09 2013-07-11 International Business Machines Corporation Managing workload distribution among computing systems to optimize heat dissipation by computing systems
US9459633B2 (en) * 2012-01-09 2016-10-04 International Business Machines Corporation Managing workload distribution among computing systems to optimize heat dissipation by computing systems
US20150169016A1 (en) * 2012-06-29 2015-06-18 Qualcomm Incorporated System and method for adaptive thermal management in a portable computing device
US9360907B2 (en) * 2012-06-29 2016-06-07 Qualcomm Incorporated System and method for adaptive thermal management in a portable computing device
US8996902B2 (en) 2012-10-23 2015-03-31 Qualcomm Incorporated Modal workload scheduling in a heterogeneous multi-processor system on a chip
CN104049714A (en) * 2013-03-15 2014-09-17 英特尔公司 Processor having frequency of operation information for guaranteed operation under high temperature events
US20140281445A1 (en) * 2013-03-15 2014-09-18 Ankush Varma Processor having frequency of operation information for guaranteed operation under high temperature events
GB2512745B (en) * 2013-03-15 2016-07-20 Intel Corp Processor having frequency of operation information for guaranteed operation under high temperature events
GB2512745A (en) * 2013-03-15 2014-10-08 Intel Corp Processor having frequency of operation information for guaranteed operation under high temperature events
US9494996B2 (en) * 2013-03-15 2016-11-15 Intel Corporation Processor having frequency of operation information for guaranteed operation under high temperature events
EP2796961A3 (en) * 2013-04-25 2016-10-26 Intel Corporation Controlling power and performance in a system agent of a processor
US9485168B2 (en) * 2013-10-03 2016-11-01 International Business Machines Corporation Temperature sensitive routing of data in a computer system
US9544221B2 (en) * 2013-10-03 2017-01-10 International Business Machines Corporation Temperature sensitive routing of data in a computer system
US20150098476A1 (en) * 2013-10-03 2015-04-09 International Business Machines Corporation Temperature sensitive routing of data in a computer system
US9838300B2 (en) 2013-10-03 2017-12-05 International Business Machines Corporation Temperature sensitive routing of data in a computer system
US20150100681A1 (en) * 2013-10-03 2015-04-09 International Business Machines Corporation Temperature sensitive routing of data in a computer system
US10114784B2 (en) 2014-04-25 2018-10-30 Liqid Inc. Statistical power handling in a scalable storage system
US9678910B2 (en) 2014-04-25 2017-06-13 Liqid Inc. Power handling in a scalable storage system
US10467166B2 (en) 2014-04-25 2019-11-05 Liqid Inc. Stacked-device peripheral storage card
US11816054B2 (en) 2014-04-25 2023-11-14 Liqid Inc. Scalable communication switch system
US10733130B2 (en) 2014-04-25 2020-08-04 Liqid Inc. Scalable storage system
US10474608B2 (en) 2014-04-25 2019-11-12 Liqid Inc. Stacked-device peripheral storage card
US12086089B2 (en) 2014-04-25 2024-09-10 Liqid Inc. Processor-endpoint isolation in communication switch coupled computing system
US11269798B2 (en) 2014-04-25 2022-03-08 Liqid Inc. Scalable communication fabric system
US10037296B2 (en) 2014-04-25 2018-07-31 Liqid Inc. Power handling in a scalable storage system
US10983941B2 (en) 2014-04-25 2021-04-20 Liqid Inc. Stacked storage drives in storage apparatuses
US9671767B2 (en) * 2014-05-14 2017-06-06 Advanced Micro Devices, Inc. Hybrid system and method for determining performance levels based on thermal conditions within a processor
US20150331433A1 (en) * 2014-05-14 2015-11-19 Advanced Micro Devices, Inc. Hybrid system and method for determining performance levels based on thermal conditions within a processor
US9652019B2 (en) 2014-06-02 2017-05-16 Advanced Micro Devices, Inc. System and method for adjusting processor performance based on platform and ambient thermal conditions
US9798636B2 (en) 2014-06-23 2017-10-24 Liqid Inc. Front end traffic handling in modular switched fabric based data storage systems
US10223315B2 (en) 2014-06-23 2019-03-05 Liqid Inc. Front end traffic handling in modular switched fabric based data storage systems
US9645902B2 (en) 2014-06-23 2017-05-09 Liqid Inc. Modular switched fabric for data storage systems
US10754742B2 (en) 2014-06-23 2020-08-25 Liqid Inc. Network failover handling in computing systems
US9684575B2 (en) 2014-06-23 2017-06-20 Liqid Inc. Failover handling in modular switched fabric for data storage systems
US10180889B2 (en) 2014-06-23 2019-01-15 Liqid Inc. Network failover handling in modular switched fabric based data storage systems
US10503618B2 (en) 2014-06-23 2019-12-10 Liqid Inc. Modular switched fabric for data storage systems
US10496504B2 (en) 2014-06-23 2019-12-03 Liqid Inc. Failover handling in modular switched fabric for data storage systems
CN105334939A (en) * 2014-06-30 2016-02-17 北京安兔兔科技有限公司 Terminal power consumption control method and terminal power consumption control device
US10001819B2 (en) 2014-09-04 2018-06-19 Liqid Inc. Dual-sided rackmount modular data processing assembly
US10362107B2 (en) 2014-09-04 2019-07-23 Liqid Inc. Synchronization of storage transactions in clustered storage systems
US10198183B2 (en) 2015-02-06 2019-02-05 Liqid Inc. Tunneling of storage operations between storage nodes
US10585609B2 (en) 2015-02-06 2020-03-10 Liqid Inc. Transfer of storage operations between processors
US9785208B2 (en) * 2015-03-20 2017-10-10 Dell Products Lp Systems and methods of adaptive thermal control for information handling systems
US10353446B2 (en) 2015-03-20 2019-07-16 Dell Products L.P. Systems and methods of adaptive thermal control for information handling systems
US20160274629A1 (en) * 2015-03-20 2016-09-22 Dominick Adam Lovicott Systems And Methods Of Adaptive Thermal Control For Information Handling Systems
US10019388B2 (en) 2015-04-28 2018-07-10 Liqid Inc. Enhanced initialization for data storage assemblies
US10191691B2 (en) 2015-04-28 2019-01-29 Liqid Inc. Front-end quality of service differentiation in storage system operations
US10402197B2 (en) 2015-04-28 2019-09-03 Liqid Inc. Kernel thread network stack buffering
US10423547B2 (en) 2015-04-28 2019-09-24 Liqid Inc. Initialization of modular data storage assemblies
US10740034B2 (en) 2015-04-28 2020-08-11 Liqid Inc. Front-end quality of service differentiation in data systems
US10108422B2 (en) 2015-04-28 2018-10-23 Liqid Inc. Multi-thread network stack buffering of data frames
US10521394B2 (en) * 2015-07-16 2019-12-31 Motorola Mobility Llc Apparatus and method for improving application performance within a group of communication devices
US20170017608A1 (en) * 2015-07-16 2017-01-19 Motorola Mobility Llc Apparatus and method for improving application performance within a group of communication devices
US20180203494A1 (en) * 2015-07-29 2018-07-19 Zte Corporation Method and apparatus for controlling hot plug operation of cpu in mobile terminal
US10599195B2 (en) * 2015-07-29 2020-03-24 Zte Corporation Method and apparatus for controlling hot plug operation of CPU in mobile terminal
US10001800B1 (en) 2015-09-10 2018-06-19 Apple Inc. Systems and methods for determining temperatures of integrated circuits
US10990553B2 (en) 2016-01-29 2021-04-27 Liqid Inc. Enhanced SSD storage device form factors
US10255215B2 (en) 2016-01-29 2019-04-09 Liqid Inc. Enhanced PCIe storage device form factors
CN107153592A (en) * 2016-03-02 2017-09-12 神讯电脑(昆山)有限公司 Electronic installation and its power management method
US11399720B2 (en) 2016-04-05 2022-08-02 Qulacomm Incorporated Circuits and methods providing temperature mitigation for computing devices
US10402363B2 (en) 2016-06-10 2019-09-03 Liqid Inc. Multi-port interposer architectures in data storage systems
US10866923B2 (en) 2016-06-10 2020-12-15 Liqid Inc. Storage control interposers in data storage systems
US11922218B2 (en) 2016-08-12 2024-03-05 Liqid Inc. Communication fabric coupled compute units
US11294839B2 (en) 2016-08-12 2022-04-05 Liqid Inc. Emulated telemetry interfaces for fabric-coupled computing units
US11880326B2 (en) 2016-08-12 2024-01-23 Liqid Inc. Emulated telemetry interfaces for computing units
US10642659B2 (en) 2016-08-12 2020-05-05 Liqid Inc. Telemetry handling for disaggregated fabric-switched computing units
US10983834B2 (en) 2016-08-12 2021-04-20 Liqid Inc. Communication fabric coupled compute units
US10592291B2 (en) 2016-08-12 2020-03-17 Liqid Inc. Disaggregated fabric-switched computing platform
US10691647B2 (en) * 2016-08-16 2020-06-23 International Business Machines Corporation Distributed file system metering and hardware resource usage
US20190034241A1 (en) * 2016-08-16 2019-01-31 International Business Machines Corporation Optimized resource metering in a multi tenanted distributed file system
WO2018041530A3 (en) * 2016-09-05 2018-06-14 Intel IP Corporation Method and device for thermal management control of an electronic device
EP3291048A1 (en) * 2016-09-05 2018-03-07 Intel IP Corporation Method and device for thermal management control of an electronic device
TWI652569B (en) 2017-01-05 2019-03-01 新唐科技股份有限公司 Apparatus, system, method and computer program product for improved regulation of electrical device's consumption of power from external power supply
US10175748B2 (en) * 2017-01-05 2019-01-08 Nuvoton Technology Corporation System, method and computer program product for improved regulation of an electrical device's consumption of power from an external power supply
US10614022B2 (en) 2017-04-27 2020-04-07 Liqid Inc. PCIe fabric connectivity expansion card
US11615044B2 (en) 2017-05-08 2023-03-28 Liqid Inc. Graphics processing unit peer-to-peer arrangements
US10795842B2 (en) 2017-05-08 2020-10-06 Liqid Inc. Fabric switched graphics modules within storage enclosures
US12038859B2 (en) 2017-05-08 2024-07-16 Liqid Inc. Peer-to-peer arrangements among endpoint devices
US10628363B2 (en) 2017-05-08 2020-04-21 Liqid Inc. Peer-to-peer communication for graphics processing units
US11314677B2 (en) 2017-05-08 2022-04-26 Liqid Inc. Peer-to-peer device arrangements in communication fabrics
US10180924B2 (en) 2017-05-08 2019-01-15 Liqid Inc. Peer-to-peer communication for graphics processing units
US10936520B2 (en) 2017-05-08 2021-03-02 Liqid Inc. Interfaces for peer-to-peer graphics processing unit arrangements
US11086371B2 (en) * 2018-06-06 2021-08-10 Lenovo (Singapore) Pte. Ltd. Heat suppression in an information processing device in an initial setting mode
US10993345B2 (en) 2018-08-03 2021-04-27 Liqid Inc. Peripheral storage card with offset slot alignment
US10660228B2 (en) 2018-08-03 2020-05-19 Liqid Inc. Peripheral storage card with offset slot alignment
US11609873B2 (en) 2019-02-05 2023-03-21 Liqid Inc. PCIe device peer-to-peer communications
US11119957B2 (en) 2019-02-05 2021-09-14 Liqid Inc. PCIe device peer-to-peer communications
US11921659B2 (en) 2019-02-05 2024-03-05 Liqid Inc. Peer-to-peer communications among communication fabric coupled endpoint devices
US10585827B1 (en) 2019-02-05 2020-03-10 Liqid Inc. PCIe fabric enabled peer-to-peer communications
US11256649B2 (en) 2019-04-25 2022-02-22 Liqid Inc. Machine templates for predetermined compute units
US11949559B2 (en) 2019-04-25 2024-04-02 Liqid Inc. Composed computing systems with converged and disaggregated component pool
US11973650B2 (en) 2019-04-25 2024-04-30 Liqid Inc. Multi-protocol communication fabric control
US11265219B2 (en) 2019-04-25 2022-03-01 Liqid Inc. Composed computing systems with converged and disaggregated component pool
US12056077B2 (en) 2019-04-25 2024-08-06 Liqid Inc. Machine templates for compute units
US11442776B2 (en) 2020-12-11 2022-09-13 Liqid Inc. Execution job compute unit composition in computing clusters
WO2024183598A1 (en) * 2023-03-03 2024-09-12 华为技术有限公司 Frequency control method for processor, and electronic device

Also Published As

Publication number Publication date
US6889332B2 (en) 2005-05-03

Similar Documents

Publication Publication Date Title
US6889332B2 (en) Variable maximum die temperature based on performance state
US7647513B2 (en) Method and apparatus for improving responsiveness of a power management system in a computing device
US6845456B1 (en) CPU utilization measurement techniques for use in power management
EP0712064B1 (en) Variable frequency clock control for microprocessor-based computer systems
EP2485117B1 (en) Method and apparatus for adjusting the voltage and frequency to minimize power dissipation in a multiprocessor system
US6795927B1 (en) Power state resynchronization
US8095818B2 (en) Method and apparatus for on-demand power management
US8122269B2 (en) Regulating power consumption in a multi-core processor by dynamically distributing power and processing requests by a managing core to a configuration of processing cores
EP1361501B1 (en) Power conservation and thermal management arrangements for computers
US8412962B2 (en) Microprocessor with improved thermal monitoring and protection mechanism
US7698583B2 (en) Microprocessor capable of dynamically reducing its power consumption in response to varying operating temperature
US7730334B2 (en) Method and apparatus for on-demand power management
US7774627B2 (en) Microprocessor capable of dynamically increasing its performance in response to varying operating temperature
JP4125370B2 (en) Method and apparatus for enhancing processor performance
US6996730B2 (en) Adjusting voltage supplied to a processor in response to clock frequency
US7770042B2 (en) Microprocessor with improved performance during P-state transitions
US20060174146A1 (en) Microprocessor performance mode control utilizing sensed temperature as an indication of microprocessor utilization
US9785218B2 (en) Performance state selection for low activity scenarios
EP1965286B1 (en) Microprocessor with improved thermal monitoring and protection mechanism
JP2001337736A (en) Computer system

Legal Events

Date Code Title Description
AS Assignment

Owner name: ADVANCED MICRO DEVICES, INC., CALIFORNIA

Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:HELMS, FRANK P.;BRINKLEY, JEFFREY A.;REEL/FRAME:012399/0224;SIGNING DATES FROM 20011203 TO 20011207

FEPP Fee payment procedure

Free format text: PAYOR NUMBER ASSIGNED (ORIGINAL EVENT CODE: ASPN); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY

STCF Information on status: patent grant

Free format text: PATENTED CASE

FPAY Fee payment

Year of fee payment: 4

FPAY Fee payment

Year of fee payment: 8

IPR Aia trial proceeding filed before the patent and appeal board: inter partes review

Free format text: TRIAL NO: IPR2015-00323

Opponent name: LG ELECTRONICS USA, INC.,LG ELECTRONICS MOBILECOMM

Effective date: 20141210

FPAY Fee payment

Year of fee payment: 12